CN1273931C - Method and system for improving image quality - Google Patents

Abstract

The invention concerns a method and a system for modifying the quality of an image derived from or addressed to a set of appliances (APP1 to n). It uses formatted data (15.1 to 15.n) related to the defects of the appliances of the set of appliances. The invention comprises: listing up sources of formatted data concerning said appliances; automatically finding said formatted data (15.1 to 15.n) associated with the appliances of said set of appliances; automatically modifying said image (1) using image processing software and/or image processing components (23.1 to 23.n) integrating said formatted data (15.1 to 15.n) thus obtained. The invention further concerns various alternatives such as modifying the formatted data on the basis of variable characteristics of the image to be processed and/or of the set of appliances. The invention enables to process the images from appliances which can be of different origins and have been commercialised gradually over time. The invention is applicable to the processing of photographic or video images, in optical devices, industrial controls, robotics, metrology and the like.

Description

Method and system for improving image quality

Technical field

The invention relates to a method and a system for improving the quality of at least one image originating from or destined for a chain of devices.

plan

method

The invention relates to a method for improving the quality of at least one image originating from or destined for a given chain of devices. The designated device chain includes at least one image capture device and/or at least one image restoration device. The image capture device and/or image restoration device has gradually been commercialized by different economic actors, belonging to an indeterminate group of devices. Defects exhibited by devices in the set of devices may be characterized by the formatted information. For the image currently under consideration, the method includes the following steps:

the step of editing a directory of sources of formatted information related to the set of devices,

the step of automatically retrieving, among the formatted information compiled in this manner, specific formatted information associated with the specified device chain,

the step of automatically modifying an image by means of image processing software and/or image processing components taking into account the specific formatting information obtained in this way,

Preferably, according to the invention, the method enables automatic retrieval by means of indicators obtained by direct or indirect analysis of:

images, and/or

devices of the device chain, and/or

means for loading images into image processing software or components, and/or

Means for loading images modified by image processing software or components into the restoration means.

Preferably, according to the invention, the devices of the device chain are identified by identifiers, in particular barcodes. The analysis performed for the purpose of retrieving the format-specific information includes the step of determining an identifier.

Preferably, according to the invention, the method is such that the image, indicator and/or identifier are contained in the same file. The combination of several technical features makes it possible to empirically exploit the method according to the invention in cases where certain means of the chain have been commercialized before the formatted information related to them has been established.

Preferably, according to the present invention, the method enables the image and at least a part of specific format information to be included in the same image file. A combination of several technical features leads to the possibility of automatically retrieving formatting information in image files.

Preferably, according to the present invention, the method further includes the step of storing at least part of the formatted information in a database in advance. The method also includes the step of updating the database.

Preferably, according to the invention, the method enables the devices of the chain of devices to have at least one variable characteristic dependent on the image, in particular focal length. A portion of the specific formatting information is related to a defect of the device having the variable feature. The method also includes the steps of:

the step of determining values for variable characteristics of said image,

A step of determining a part of said specific formatting information by taking into account variable characteristic values obtained in this way.

Thus, applying the method to a device with variable features is equivalent to applying the method to a device without any variable features.

Preferably, according to the invention, the images are contained in a file. The method makes it possible to use, for determining the value of this variable characteristic, data present in the file, in particular data such as the focal length, in a format such as the Exif standard. The combination of technical features leads to the possibility of empirically using the method according to the invention in the case of devices with variable features that have been commercialized before the formatted information associated therewith has been established.

Preferably, according to the invention, the method is such that, for improving the quality of at least one image originating from or destined for a chain of devices:

A virtual device is determined which exhibits a defect corresponding to at least a partial defect of at least one device of the device chain, said at least partial defect hereinafter referred to as an original defect.

determining virtual formatting information associated with the defect of the virtual device,

Specific formatting information associated with the group of devices of the device chain is determined, and the virtual formatting information is used in place of specific formatting information associated with the original defect.

The combination of technical features can lead, in this way, to use the obtained formatting information more simply, and to calculate the modifications to the image faster by using less memory and or with higher precision.

Preferably, according to the invention, the method is designed to improve the quality of at least one color plane of a color image. The color plane is characterized by a specified color. The specific formatting information also includes data related to the designated color. To modify the image, a color plane is calculated using the data associated with the specified color and image.

Preferably, according to the invention, the method further comprises the step of calculating unknown formatting information in case the process of retrieving specific formatting information for one of the devices of the chain of devices is unsuccessful.

Preferably, according to the present invention, the method further comprises:

by measuring the device defect, and/or

By simulating the device

The step of computing unknown formatted information associated with devices of the device chain.

Preferably, according to the present invention, the method further comprises, for the image capture devices of the device chain, the step of calculating unknown formatting information:

constructing a synthetic image class by specifying mathematical projections from at least one reference scene onto a surface,

by capturing at least one reference image of each reference scene by means of the image capture device,

Within a set of parameterizable transformation models, the transformation model is selected such that the reference image can be transformed into a transformed image that is close to the synthetic image class of the reference scene.

The transformed image exhibits deviations from the synthetic image class. The unknown formatting information consists at least in part of parameters of the selected parameterizable transformation model.

Preferably, according to the present invention, the method also includes:

the step of calculating the deviation between the class of the transformed image and the synthesized image,

The step of associating the deviation with said unknown formatting information.

The combination of technical features makes it possible to derive three-dimensional normalized information about the scene. The combination of technical features makes it possible to combine multiple images obtained by multiple image capture devices that have undergone the same formatting process.

Preferably, according to the invention, the method enables a device in the chain of devices to have at least one variable characteristic dependent on the image, in particular focal length and/or aperture. Part of this specific formatted information is related to the defects of the device with this/these variable characteristics. Each variable feature can be associated with a value to form a combination comprising the set of variable features and values. The method also includes the step of determining some unknown formatting information:

By selecting a predetermined combination,

For each predetermined combination, by employing the process of repeating the preceding steps of the method,

From the unknown formatting information obtained at the end of the iterative process, by employing a process of interpolating the unknown formatting information associated with an arbitrary combination.

Preferably, according to the invention, the method further comprises, for an image restoration device of the chain of devices, the step of generating data characterizing defects of the image restoration device, in particular distortion characteristics. The unknown formatting information consists at least in part of data characterizing a defect of the recovery device.

Preferably, according to the invention, the method is such that specific formatting information relating to one or more devices of the chain of devices is determined in such a way that it can be applied to similar devices. The combination of technical features is such that only a limited amount of formatting information is required for the method employed.

Preferably, according to the invention, the method is such that the image includes relevant information, in particular a digital signal. The steps of the method are employed in such a way that they retain or modify relevant information.

Preferably, according to the invention, the method further comprises the step of associating information with the modified image, in particular information indicating that it has been modified.

Preferably, according to the invention, the method is more particularly designed to improve the visual quality of the image for a viewer. The formatted information related to the defects of the devices of the device chain comprises formatted information related to the visual characteristics of the observer, in particular the visual characteristics of the observer's eyes and/or brain.

Application

The invention also relates to the use of the above method. The purpose of the application is to improve the quality of the images processed by the image processing software or image processing components by correcting the effect of at least one defect of the devices of the device chain. The combination of technical features makes it possible to improve, if not perfect, the quality of the processed image without relying on expensive equipment.

Preferably, the purpose of the application is that the quality of the images processed by the image processing software or image processing components is comparable to the quality of the images produced with the reference device chain.

Preferably, the application is such that, for a quality of the processed image comparable to that produced using the reference device chain, the formatted information associated with the reference device chain is generated taking into account imperfections of the reference device chain.

system

The present invention relates to a system for improving the quality of at least one image originating from or destined for a given chain of devices. The specified chain of devices includes at least one image capture device and/or at least one image restoration device. The image capture device and/or image restoration device, which has been gradually commercialized by different economic players, belongs to an undecided group of devices. Defects possessed by devices in the set of devices may be characterized by formatting information. For the image in question, the system includes data processing means capable of:

edit the catalog of sources of formatted information associated with devices in the group of devices,

automatically retrieve the specific formatting information associated with the specified device chain among the formatting information edited in this manner,

automatic modification of images by means of image processing software and/or image processing components taking into account the specified formatting information obtained in this way,

Preferably, according to the present invention, the system enables the data processing device to perform automatic retrieval by means of an index obtained by directly or indirectly analyzing the following:

images, and/or

devices of the device chain, and/or

means for loading images into image processing software or components, and/or

Means for loading images modified by image processing software or components into the restoration means.

Preferably, according to the invention, the devices of the device chain are identified by identifiers, in particular barcodes. The analyzing means comprise means for determining the identifier.

Preferably, according to the invention, the system is such that the image, index and/or identifier are contained in the same file.

Preferably, according to the invention, the system causes the image and at least a portion of specific formatting information to be included in the same image file.

Preferably, according to the present invention, the system further includes storage means for storing at least a part of the formatting information in the database in advance. The system also includes updating means for updating the database.

Preferably, according to the invention, the system is such that the devices of the chain of devices have at least one variable characteristic dependent on the image, in particular focal length. A portion of the specific formatting information is related to a defect of the device having the variable feature. The system also includes computing means for determining:

the value of the variable feature for the image in question,

Part of the specific formatting information, by taking into account variable values obtained in this way.

Preferably, according to the invention, the image is contained in a file. The system is such that, in order to determine the value of the variable characteristic, the system comprises data processing means for processing the data present in the file, in particular such as the focal length, in a format such as the Exif standard.

Preferably, according to the invention, the system is such that in order to improve the quality of at least one image originating from or destined for a chain of devices, the system comprises data processing means to determine:

a virtual device exhibiting a defect corresponding to at least a portion of the defect of at least one device of the chain of devices, said at least a portion of the defect being referred to hereinafter as the original defect,

Virtual formatting information associated with the defect of this virtual device.

The system is such that, in order to determine specific formatting information associated with the group of devices of the device chain, the data processing device comprises replacement means for replacing specific formatting information associated with the original defect with virtual formatting information. information.

Preferably, according to the invention, the system is designed to improve the quality of at least one color plane of a color image. The color plane is characterized by a specified color. The specific formatting information also includes data related to the designated color. The system includes computing means for computing a color plane using data associated with the specified color and image.

Preferably, according to the invention, the system further comprises computing means for computing unknown formatting information in case the process of retrieving specific formatting information for one of the devices of the chain of devices is unsuccessful.

Preferably, according to the invention, the system is such that the computing means for computing unknown formatted information relating to a device of the chain of devices comprises processing means for measuring defects and/or simulating the device.

Preferably, according to the invention, the method further comprises, for the image capture means of the chain of means, a class of synthetic images constructed by a specified mathematical projection of at least one reference scene onto a surface, for computing the unknown A computing device that formats information. The image capture device captures at least one reference image of each reference scene. The unknown formatting information is computed by selecting, within a set of parameterizable transformation models, a transformation model that enables the transformation of the reference image into a transformed image that approximates the class of the synthetic image of the reference scene. The transformed image exhibits deviations compared to the synthetic image class. The unknown formatting information is formed at least in part by parameters of the selected parameterizable transformation model.

Preferably, according to the present invention, the system further includes a data processing device for

Calculate the deviation between the transformed image and the synthesized image class,

Associate this deviation with unknown formatting information.

Preferably, according to the invention, the system is such that a device in the chain of devices has at least one variable characteristic dependent on the image, in particular focal length and or aperture. Part of this specific formatted information is related to the defects of the device with this/these variable characteristics. Each variable feature can be associated with a value to form a combination comprising the set of variable features and values. The system also includes data processing means for determining a portion of the unknown formatted information:

By selecting a predetermined combination,

For each predetermined combination, by employing an iterative process for the computing means and data processing means described above,

A process of interpolating the unknown formatting information associated with an arbitrary combination is employed based on the unknown formatting information obtained at the end of the iterative process.

Preferably, according to the invention, the system further comprises, for the image restoration means of the chain of means, data processing means for generating data characterizing defects, in particular distortion signatures, of the image restoration means. The unknown formatting information consists at least in part of data characterizing a defect of the recovery device.

Preferably, according to the invention, the system is such that specific formatting information relating to one or more devices of the device chain is determined in such a way that it can be applied to similar devices.

Preferably, according to the invention, the system causes the image to include relevant information, especially digital signals. Use of this system enables the retention or modification of relevant information.

Preferably, according to the invention, the system further comprises data processing means for associating information with the modified image, in particular information indicating that it has been modified.

Preferably, according to an alternative embodiment of the invention, the system is more specifically designed to improve the visual quality of the image for the observer. The formatted information related to defects of the devices of the device chain further comprises formatted information related to the visual characteristics of the observer, in particular abnormalities of the eyes and/or brain of the observer.

A detailed description

Other characteristics and advantages of the present invention will become apparent by reading the description of the non-limitative application examples provided and the accompanying drawings of the present invention.

Figure 1 shows a schematic diagram of image capture,

Fig. 2 shows the schematic diagram of image restoration,

Figure 3 shows a schematic diagram of an image pixel,

Figures 4a and 4b show two schematic diagrams of a reference scene,

Figure 5 shows an organizational diagram of a method that can be used to calculate the difference between a mathematical image and a corrected image,

Figure 6 shows an organizational diagram of a method that can be used to obtain an optimal restoration transformation for an image restoration device,

Figure 7 shows a schematic diagram of the elements that make up the system to which the present invention is applied,

Fig. 8 shows a schematic diagram of fields of formatted information,

Figure 9a shows a schematic front view of a mathematical point,

Figure 9b shows a schematic front view of a real point of an image,

Figure 9c shows a simple side view of a mathematical point,

Figure 9d shows a schematic cross-sectional view of a real point of an image,

Figure 10 shows a schematic diagram of a feature point array,

Figure 11 shows an organizational diagram of methods that can be used to obtain formatted information,

Figure 12 shows an organizational diagram that can be used to obtain an optimal transformation method for an image capture device,

Fig. 13a-13c shows the connection diagram that can be used for the application example of the system of correcting image, and wherein OU1, 2 represent user's operator 1, 2, BD represents database, UC represents central unit,

Fig. 14a-14c shows the organizational diagram of the application example of the method for applying automatic image correction,

Figure 15 shows an organizational diagram of a method that can be used to replace a chain of devices with a virtual device,

Figure 16.1 shows a schematic diagram of a device with a defect,

Figure 16.2 shows a schematic diagram of a device with variable characteristics,

Figure 16.3 shows a schematic diagram including one or more visual defects of the observer,

Figure 16.4 shows a schematic diagram of processing the characteristics of a virtual device,

Figure 16.5 shows a schematic diagram of adding information associated with a corrected image,

Figure 16.6 shows a schematic diagram showing how formatted information can be associated with one or more devices,

Figure 17 shows a description of an example using the method and system according to the present invention,

Figure 18 shows a depiction of an example using the method and system according to the invention in the case of a color image.

Device Chain

Specifically referring to FIG. 17 , the concept of the device chain P3 is now described. Devices of the device chain, especially image capture devices and/or image restoration devices, are gradually commercialized by different economic actors and they belong to an undefined group of devices, also defined as device group P75.

Within the meaning of the present invention, the device may in particular be:

Image capture devices, such as disposable camera devices, digital camera devices, reflective devices, scanners, facsimile machines, endoscopes, camcorders, surveillance cameras, game consoles, integrated into or connected to telephones, personal digital assistants or computer cameras, thermal cameras, or echographic devices,

An image restoration device or image restoration device 19, such as a screen, a projector, a television, a virtual reality eyepiece or a printer,

a device, including its installation, such as projectors, screens and how they are positioned,

the position of the observer relative to the image restoration device which specifically introduces parallax,

people with visual impairments such as astigmatism,

A device which is expected to be imitated for producing images similar in appearance to those produced by, for example, Lycra devices,

Image processing equipment with edge effects that add blur, such as zoom software,

a virtual device equivalent to multiple devices,

More complex installations such as all-in-one scanner/fax/printer, photo-printing mini-lab, video conferencing installation may be considered as one or more installations.

A device chain P3 is defined as a group of devices. The concept of sequence is also included in the concept of device chain P3.

The following instances make up the device chain P3:

single device,

Image capture device and image restoration device,

For example photographic devices, scanners or printers in photo-printing microlabs,

such as a digital camera or printer in a photo-printing microlab,

such as a scanner, screen or printer in a computer,

screens or projectors, and the human eye,

A device and another device that is expected to be imitated,

photographic devices and scanners,

image capture device and image processing software,

Image processing software and image restoration device,

Combinations of the above examples,

Another set of devices.

defect

With specific reference to Figure 17, the concept of defect P5 will now be described. Defects P5 of device group P75 are defined as defects involving features of the optical system and/or sensors and/or electronic units and/or software integrated into the device; examples of defects P5 include: distortion, blurring, vignetting, chromatic aberration, chromatic aberration Rendering of color, flare uniformity, sensor noise, graininess, astigmatism and spherical aberration.

Image

Referring specifically to Fig. 17, the concept of picture P2 will now be described. Image P2 is defined as a digital image acquired or modified or restored by the device. Image P2 may originate from a device in device chain P3. Image P2 may be sent to a device in device chain P3. More generally, image P2 may come from and/or be sent to device chain P3. In an animated picture composed of still pictures arranged in time series, such as a video picture, picture P2 is defined as a still picture in the picture sequence.

Formatting information

Referring specifically to Figure 17, the concept of formatting information 15 will now be described. The formatted information 15 is defined as data relating to the defect P5 of one or more devices P25 of the device chain P3, which data enables the image processing device to improve the quality of the image P2 by assessing the defect P5 of the device. To generate the formatted information 15, different methods and systems can be used based on the capture or recovery of measurements and/or references, and/or simulations.

To generate formatted information 15 it is possible to use the method and system described, for example, in the International patent application filed on the same day as the present application in the name of Vision IQ, entitled "Method and system for producing formatted information related to the defects of appliances of an appliance chain and formatted information addressed to image-processing means". This application describes a method for generating formatted information 15 related to a defect P5 of a device in a device chain P3. Formatting information 15 is sent to image processing means, in particular software, for the purpose of improving the quality of the images processed by the image processing means. The device chain P3 consists in particular of at least one image capture device and/or at least one image restoration device and/or at least one observer. This method comprises a step for generating data characterizing defects P5 of the devices in the chain P3 of devices. This data is formatting information 15 .

For generating formatted information 15 it is possible to use methods and systems such as those described in the International patent application filed on the same date as the present application in the name of Vision IQ, entitled: "Method and system for producing formatted information related to geometric distortions ". This application describes a method for generating formatted information 15 related to devices in a device chain 3 . The device chain P3 consists in particular of at least one image capture device and/or at least one image restoration device. The method comprises the step of generating formatted information 15 related to the geometric distortion of at least one device in the chain.

Preferably, the device has the possibility to capture or restore the image on the medium. According to picture (I), the device contains at least one fixed feature and/or one variable feature. A fixed characteristic and/or a variable characteristic can be associated with one or more characteristic values, in particular focal length and/or focusing characteristics and characteristic values of related characteristics. The method includes the step of generating precise formatted information related to geometric distortion of the device from a precise field. This formatting information 15 may contain precise formatting information.

For generating formatted information 15 it is possible to use methods and systems such as described in the International Patent Application filed on the same date as the present application in the name of Vision IQ, entitled: "Method and system for producing formatted information related to the defects of at least one appliance of a chain, especially to blurring". This application describes a method for generating formatted information 15 related to devices of a device chain P3. The device chain P3 consists in particular of at least one image capture device and/or at least one image restoration device. The method comprises the step of generating formatted information 15 relating to a defect P5 of at least one device of the chain. Preferably, depending on the image (I), the means operable to capture or restore the image comprise at least one fixed feature and/or one variable feature. The fixed and/or variable characteristic can be associated with one or more characteristic values, in particular focal length and/or focusing characteristics and characteristic values of related characteristics. The method comprises the step of generating, from a precise field, precise formatting information relating to the defect P5 of the device. This formatting information 15 may contain precise formatting information.

For providing formatted information 15 it is possible to use methods and systems such as those described in the international patent application filed on the same date as the present application in the name of Vision IQ, entitled: "Method and system for providing formatted information in a standard format to image-processing means". This application describes a method for providing formatted information 15 in a standardized format to image processing devices, especially software and/or components. The formatting information 15 is related to a defect P5 of a device chain P3. This device chain P3 consists in particular of at least one image capture device and/or at least one image restoration device. The image processing device uses the formatting information 15 to improve the quality of at least one image P2 originating from or destined for the device chain P3. The formatting information 15 comprises data characterizing defects P5 of the image capture device, in particular distortions, and/or data characterizing defects of the image restoration device, in particular distortions.

The method comprises the step of filling at least one field of a standard format with formatting information 15 . The field is specified by its field name, and the field includes at least one field value.

To generate the formatted information 15 it is possible to use, for example, the method and system described in the International patent application filed on the same date as the present application in the name of Vision IQ, entitled: "Method and system for reducing update frequency of image processing means" . This application describes a method for reducing the update frequency of an image processing device, in particular of software and/or components. The image processing means make it possible to vary the quality of the digital images originating from or sent to the chain of means P3. The device chain P3 consists in particular of at least one image capture device and/or at least one image restoration device. The image processing device uses formatted information 15 related to a defect P5 of at least one device in the device chain P3. The formatting information 15 depends on at least one variable. Formatting information makes it possible to establish correspondence between a part of variables and a part of identifiers. By means of the identifier, it is possible to determine the value of the variable corresponding to the identifier by considering the identifier and the image. The combination of technical features makes it possible to determine the value of a variable, especially if the physical significance and/or content of this variable is only available after the distribution of the image processing means. The combination of technical features also makes it possible to spatially separate the time between two updates of the correction software. The combination of technical features also makes it possible for several economic actors of production installations and/or image processing installations to update their products independently of other economic actors, even if the latter radically changes the characteristics of the product or cannot force customers to update their products. The combination of technical features also makes it possible, starting with a limited number of economic actors and advanced users, to gradually start using new functions.

To use formatted information 15, it is possible to use, for example, the method and system described in the international patent application filed on the same date as the present application in the name of Vision IQ, entitled: "Method and system for calculating a transformed image from a digital image and formatted information related to a geometric transformation". This application describes a method for calculating a transformed image from a digital image and formatting information 15 related to the geometric transformation, in particular formatting information 15 concerning distortion and/or color difference of the device chain P3. The method includes the step of computing a transformed image from an approximation of the geometric transformation. It follows that such computing is economical in terms of storage resources, storage bandpass, computing power and therefore power consumption. It also follows that the converted image is free from noticeable or objectionable defects in subsequent use.

To use formatted information 15, it is possible to use, for example, the method and system described in the International Patent Application filed on the same day as this application in the name of Vision IQ, entitled: "Method and system for modifying a digital image, taking into account its noise". This application describes a method for computing a transformed image from a digital image and formatting information 15 related to a defect P5 of a chain of devices P3. The device chain P3 includes image capture devices and/or image restoration devices. The device chain P3 includes at least one device. The method includes the step of automatically determining characteristic data from the formatted information 15 and/or the digital image. The combination of technical features renders the transformed image free from noticeable or objectionable defects, in particular noise-related defects, during subsequent use.

Image processing device

The following example shows one way of generating formatted information.

1 shows a scene 3 comprising an object 107, a sensor 101 and a sensor surface 110, an optical center 111, an observation point 105 located on the sensor surface 110, an observation direction 106 passing through the observation point 105, the optical center 111 and the scene 3, and The surface 10 is geometrically associated with the sensor surface 110 .

FIG. 2 shows the image 103, the image restoration means 19, and the restored image 191 obtained on the restoration medium 190. As shown in FIG.

FIG. 3 shows a scene 3 , the image capture device 1 and an image 103 composed of pixels 104 .

Figures 4a and 4b show two alternative versions of the reference scenario 9.

Fig. 5 shows an organizational chart using: scene 3, mathematical projection 8, which provides a mathematical image 70 of scene 3, real projection 72, which provides an image 103 of scene 3 for features 74 used , a parameterizable transformation model 12 that provides a corrected image 71 of the image 103 with a difference 73 compared to the mathematical image 70 .

Figure 6 shows a histogram using: an image 103, a true restoration projection 90, which provides a restored image 191 of the image 103 for the restoration features 95 used, a parameterizable restoration transformation model 97, which A corrected restored image 94 of image 103 is provided, and a mathematically restored projection 96 is provided which provides a mathematically restored image 92 of the corrected restored image 94 with a restored difference 93 compared to restored image 191 .

FIG. 7 shows a system comprising an image capture device 1 consisting of an optical system 100 , a sensor 101 and an electronics unit 102 . Fig. 7 also shows storage area 16 containing image 103, database 22 containing formatting information 15, and means 18 for transmitting completed image 120 to computing device 17 containing image processing software 4, image 120 consists of image 103 and formatting information 15 .

FIG. 8 shows formatted information 15 made up of fields 90 .

Figures 9a to 9d show a mathematical image 70, image 103, mathematical position 40 and mathematical shape 41 of a point compared to the actual position 50 and actual shape 51 of the corresponding point in the image.

Figure 10 shows an array 80 of feature points.

FIG. 11 shows an organizational chart using the image 103, the features 74 used and the database 22 of features. This formatting information 15 is obtained from the used features 74 and stored in the database 22 . The finished image 120 is obtained from image 103 and formatting information 15 .

Fig. 12 shows an organizational chart using a reference scene 9, a mathematical projection 8, which gives the synthetic image class 7 of the reference scene 9, and a real projection 72, which gives the reference image 11 of the reference scene 9, with Feature 74 for use. The organization map also contains a parameterizable transformation model 12 which gives a transformed image 13 of the reference image 11 . The transformed image 13 shows a deviation 14 in comparison with the composite image 7 .

Definition and Detailed Description

Other features and advantages of the present invention will become apparent by reading the following:

The definitions of the technical terms used are explained below with reference to the suggestive and non-limiting examples of FIGS. 1 to 12 .

Description of Figures 1 to 12.

Scenes

A scene 3 is defined as a location in three-dimensional space containing an object 107 illuminated by a light source.

Image capture device, image, image capture

Referring to Fig. 3 and Fig. 7, what is the image capturing device 1 and the image 103 will now be described. The image capture device 1 is defined as a device consisting of an optical system 100 , one or more sensors 101 , an electronic unit 102 and a storage area 16 . By means of the image capture device 1 it is possible to capture still or animated digital images 103 from the scene 3 which are recorded in the storage area 16 or which are transmitted to an external device. The moving image is composed of a sequence of still images 103 in time. The specific form of the image capturing device 1 can be a photographic device, a video camera, a camera connected or integrated into a PC, a camera connected or integrated into a personal digital assistant, a camera connected or integrated into a phone, a video conferencing device, a measuring camera Or devices sensitive to wavelengths other than visible light, such as thermal cameras.

Image capture is defined as the method used by the image capture device 1 to compute the image 103 .

In the case of a device equipped with several interchangeable components, in particular the optical system 100, the image capture device 1 is defined as a particular configuration of the device.

    Image Restoration Device, Image Restoration, Image Restoration

Referring to Fig. 2, what is called the image restoration device 19 will now be described. The specific form of such image restoration device 19 can be a visual display screen, a TV screen, a flat screen, a projector, a virtual reality eyepiece, and a printer.

Such image recovery device 19 is made up of the following:

electronic unit,

one or more light sources, electron sources or ink sources,

Modulator(s): A device for modulating light, electrons, or ink,

Focusing equipment in the form of an optical system in the case of a light projector, an electron beam focusing coil in the case of a CRT screen, an optical filter in the case of a flat screen,

Recovery medium 190, in the specific form of: a screen in the case of a CRT screen, a flat screen or a projector, a print medium on which to print in the case of a printer, and a space in the case of a virtual image projector virtual surface in .

From an image 103 , a restored image 191 on a restoration medium 190 can be obtained by means of the image restoration device 19 .

Animated images are composed of time-wise sequences of still images.

Image restoration is defined as a method for displaying or printing an image by means of the image restoration device 19 .

In case the recovery device 19 is equipped with a plurality of interchangeable parts or parts which are movable relative to each other, in particular the recovery medium 190, the image recovery device 19 is limited to a special configuration.

          Sensor Surface, Optical Center, Focal Length

Referring to FIG. 1 , the definition of the sensor surface 110 will now be described.

The sensor surface 110 is defined as the spatial shape drawn by the sensitive surface of the sensor 101 of the image capture device 1 at the time of image capture. This surface is usually planar.

Optical center 111 is defined as a point in space associated with image 103 at the time of image capture. In case the sensor surface 110 is a plane, the focal length is defined as the distance between this point 111 and the plane 110 .

Pixel, pixel value, exposure time

Referring to Fig. 3, what a pixel 104 is and what the pixel values are will now be described.

A pixel 104 is defined as a unit area of the sensor surface 110 obtained by creating a generally regular grid on the sensor surface 110 . A pixel value is defined as a number associated with the pixel 104 .

Image capture is defined as determining the value of each pixel 104 . This set of values makes up the image 103 .

During image capture, the pixel value is obtained by integrating, over a time period defined as the exposure time, the fraction of the luminous flux obtained from the scene 3 through the optical system 100 on the surface of the pixel 104 , and convert the result of the integration into a value. The integration of the luminous flux and/or the conversion of the result of this integration into a digital value is performed by means of the electronic unit 102 .

This definition of the pixel value concept is applicable in the case of black and white or color images 103, whether they are still or animated.

However, depending on the situation, some of the luminous fluxes currently discussed are obtained in different ways:

a) In the case of a color image 103, the sensor surface 110 is generally composed of multiple types of pixels 104, each associated with a different wavelength of light flux, eg red, green and blue pixels.

b) In the case of a color image 103, there may also be several sensors 101 placed side by side, each receiving part of the light flux.

c) In the case of a color image 103, the colors used may be different from red, green and blue, eg North American NTSC TV, and they may exceed three colors in number.

d) Finally, in the case of an interlaced telescan camera, an animated picture is produced consisting alternately of pictures 103 containing even lines and pictures 103 containing odd lines.

The configuration used, the adjustment used, the feature used

The configuration used is defined as the list of movable parts of the image capture device 1 , for example: the optical system 100 , if it is interchangeable, is arranged on the image capture device 1 . The configuration used is specifically characterized by the following characteristics:

the type of optical system 100,

Serial number or any other identification of the optical system 100.

The adjustments used are limited to:

the used configuration defined above, and

Values that affect the content of the image 103, manually or automatically adjusted, available in the configuration being used. These adjustments can be made by the user, in particular by pressing a button, or can be calculated by the image capture device 1 . These adjustments may be stored in the device, especially on a removable medium, or on any device connected to the device. These adjustments may specifically include focus adjustments, adjustments to the aperture and focal length of the optical system 100, exposure time adjustments, white balance adjustments, and joint image processing adjustments such as digital zoom, compression, and contrast.

The feature 74 used or the set of features 74 used is defined as:

a) Parameters related to the inherent technical features of the image capture device 1 , which parameters are determined during the design phase of the image capture device 1 . For example, these parameters may include the formula of the optical system 100 configured in use, which affects the geometric imperfections and sharpness of the captured image; the formula of the optical system 100 configured in use specifically includes the shape, arrangement and Material.

These parameters can also include:

The geometry of the sensor 101, in other words the shape and relative arrangement of the sensor surface 110 and the pixels 104 on this surface,

noise generated by the electronic unit 102,

The equation used to convert luminous flux to pixel values.

b) parameters associated with the inherent technical characteristics of the image capture device 1, determined during the manufacturing phase of the image capture device 1, and, in particular:

the precise positioning of the lenses in the optical system 100 of the configuration used,

Precise positioning of the optical system 100 relative to the sensor 101 .

c) parameters associated with the technical characteristics of the image capture device 1, determined at the moment of capturing the image 103, and, in particular:

the position and orientation of the sensor surface 110 relative to the scene 3,

the adjustment used,

External factors, such as temperature, if they have an effect.

d) User preferences, especially the color temperature used for image restoration. For example, these preferences are selected by the user by pressing a button.

Observation point, observation direction

Referring to FIG. 1 , it will now be described what an observation point 105 and an observation direction 106 are.

Mathematical surface 10 is defined as a surface geometrically associated with sensor surface 110 . For example, if the sensor surface is planar, the math surface 10 will likely coincide with the sensor surface.

The viewing direction 106 is defined as a straight line passing through at least one point of the scene 3 and passing through the optical center 111 . The observation point 105 is defined as the intersection of the observation direction 106 and the surface 10 .

Observing color, observing intensity

Referring to Fig. 1, what is meant by observed color and observed intensity will now be described. Observation color is defined as the color of light emitted, transmitted or reflected by the scene 3 along the observation direction 106 at a given moment and observed from the observation point 105 . The observed intensity is defined as the intensity of light emitted by the scene 3 along the observed direction 106 at the same time and observed from the observed point 105 .

Color can be characterized in particular by light intensity as a function of wavelength, or else by two values measured with a colorimeter. Intensity can be represented by a value, such as that measured with a photometer.

The observed color and the observed intensity depend in particular on the relative position of the object 107 in the scene 3 , the existing illumination sources, and the transparency and reflection characteristics of the object 107 at the time of observation.

Mathematical projection, mathematical image, mathematical point, mathematical color of point,

Mathematical Strength of Point, Mathematical Shape of Point, Mathematical Position of Point

With specific reference to Figures 1, 5, 9a, 9b, 9c and 9d, the following concepts are now described: mathematical projection 8, mathematical image 70, mathematical point, mathematical color of point, mathematical intensity of point, mathematical shape 41 of point and Mathematics position 40.

Referring to FIG. 5 , it will now be described how a mathematical image 70 is constructed from specified mathematical projections 8 of at least one scene 3 on a mathematical surface 10 .

What is meant by a designated mathematical projection 8 is first described.

A specified mathematical projection 8 associates the mathematical image 70 with:

Scene 3 when capturing image 103,

and feature 74 used.

A specified mathematical projection 8 is a transformation used to characterize each point of the mathematical image 70 from the scene 3 and the features 74 used when the image was captured.

Preferably, the mathematical projection 8 is defined using the following:

The mathematical position 40 of this point is defined as the position of the observation point 105 on the mathematical surface 10 .

The point mathematical shape 41 is defined as the geometric point-like shape of the observation point 105 .

The mathematical color of the point is defined as the observed color.

The mathematical intensity of this point is defined as the observed intensity.

A math point is defined as an association of math position 40 , math shape 41 , math color and math intensity for the observation point 105 under consideration. Mathematical image 70 is constructed from the set of said mathematical points.

The mathematical projection 8 of the scene 3 is a mathematical image 70 .

real projection, real point, real color of point,

The true intensity of the point, the true shape of the point, and the true position of the point

With specific reference to Figures 3, 5, 9a, 9b, 9c and 9d, the following concepts are now described: real projection 72, real point, real color of a point, real intensity of a point, real shape 51 of a point and real position 50 of a point.

During image capture, the image capture device 1 associates the image 103 of the scene 3 with the features 74 used. Originating from scene 3 , light along observation direction 106 passes through optical system 100 and reaches sensor surface 110 .

For the observation direction, a defined real point is now obtained, which shows a difference compared to the mathematical point.

Referring to Figures 9a to 9d, the difference between real points and mathematical points will now be described.

The real shape 51 associated with said viewing direction 106 is not a point on the sensor surface, but has the shape of a cloud in three-dimensional space intersecting one or more pixels 104 in three-dimensional space. Specific causes of these differences are: coma, spherical aberration, astigmatism, grouping into pixels 104, chromatic aberration, depth of field, diffraction, parasitic reflections, and field curvature of the image capture device 1 . They give the image 103 the impression of being blurry or lacking in sharpness.

Furthermore, the real position 50 associated with the viewing direction 106 shows a difference compared to the mathematical position 40 of the point. The specific cause of this difference is geometric distortion, which creates the impression of distortion: for example, vertical walls appear curved. This difference is also caused by the fact that there are a finite number of pixels 104 and therefore real positions 50 can only have a finite number of values.

Furthermore, the real intensity associated with the viewing direction 106 shows a difference compared to the mathematical intensity of the point. The specific reasons for these differences are gamma and vignetting: for example, the edges of the image 103 appear darkened. In addition, noise may be added to the signal.

Finally, the real color associated with the viewing direction 106 exhibits a difference compared to the mathematical color of the point. The specific causes of these differences are gamma and color cast. In addition, noise may be added to the signal.

A real point is defined as the currently considered connection of real position 50 , real shape 51 , real color and real intensity for viewing direction 106 .

The real projection 72 of scene 3 consists of this set of real points.

      Parametric transformation models, parameters, corrected images

A parameterizable transformation model 12 (or simply a parameterizable transformation 12) is defined as a mathematical transformation that yields the corrected image 71 from the image 103 and the values of the parameters. As suggested below, said parameters may be calculated specifically from the features 74 used.

By means of said transformation, for each real point of image 103, it is possible to determine the correction of said real point from the value of the parameter, from the real position of said real point, and from the value of a pixel of image 103 location, the corrected color of the true point, the corrected strength of the true point, and the corrected shape of the true point. For example, the corrected position can be calculated by means of a polynomial of fixed degree as a function of the true position, the coefficients of which depend on the value of said parameter. For example, corrected color and corrected intensity may be weighted sums of pixel values, the coefficients being dependent on parameter values and true positions, or may be non-linear functions of image 103 pixel values.

The parameters may specifically include: the focal length or related values of the optical system 100 configured in use, such as the position of a group of lenses, the focusing or related values of the optical system 100 configured in use, such as the position of a group of lenses, and the optical system 100 configured in use. The aperture or related value of the system 100, such as the position of the aperture.

Difference Between Mathematical Image and Corrected Image

With reference to Fig. 5, for given scene 3 and used feature 74, the difference between mathematical image 70 and corrected image 71 is defined as: by characterizing the position, color, intensity of all or part of correction points and all or part of mathematical points and one or more values determined by the numeric value of the shape.

For example, for a given scene 3 and features 74 used, the difference 73 between the mathematical image 70 and the corrected image 71 can be determined as follows:

There are selected feature points, for example, points that may be an orthogonal array 80 of regularly placed points, as shown in FIG. 10 .

For example, for each feature point, the difference 73 is calculated by summing the absolute values of the differences between each number representing position, color, intensity and shape, respectively, for the calibration point and the mathematical point. The summation function of the absolute values of the above differences can be replaced by another function such as mean, sum of squares and any other function capable of combining these numbers.

Reference scene

Reference scene 9 is defined as scene 3 for which certain characteristics are known. For example, Fig. 4a shows a reference scene 9 consisting of a sheet of paper with regularly placed black solid circles. Figure 4b shows another sheet, with the same circles, but also with colored lines and areas. Circles are used to measure a point's true position 50, lines are used to measure a point's true shape 51, and colored areas are used to measure a point's true color and true intensity. The reference scene 9 can consist of a material other than paper.

Reference Image

Referring to Fig. 12, the concept of the reference image 11 will now be described. A reference image 11 is defined as an image of the reference scene 9 obtained using the image capture device 1 .

Composite image, composite image class

Referring to Fig. 12, the concept of composite images and composite image classes 7 will now be described. A composite image is defined as a mathematical image 70 obtained by a mathematical projection 8 of a reference scene 9 . A synthetic image class 7 is defined as a set of mathematical images 70 obtained by mathematical projection 8 to one or more reference scenes 9 for one or more sets 74 of features used. In the case of only one reference scene 9 and only one set of features 74 used, the composite image class 7 contains only one composite image.

Transform image

Referring to Fig. 12, the concept of the transform image 13 will now be described. A transformed image 13 is defined as a corrected image obtained by applying a parameterizable transformation model 12 to a reference image 11 .

Transformed image similar to composite image class, bias

Referring to Fig. 12, the concept of a transformed image 13 close to the synthetic image class 7 and the concept of a deviation 14 will now be described.

The difference between transformed image 13 and composite image class 7 is defined as the smallest difference between said transformed image 13 and any composite image of said composite image class.

In the following it will be described how to choose among a plurality of parameterizable transformation models 12 that can be used to transform each reference image 11 to the corresponding The synthetic image 7 of the reference scene 9 like 11 approximates the transformed image 13 .

Given that a given reference scene 9 is associated with a given set of features 74 used, the parameterizable transformation 12 (and its parameters) can be chosen such that it can be used to transform the reference image 11 into and synthesize Image class 7 compares the transformed image 13 showing the smallest difference. Then, the synthetic image class 7 and the transformed image 13 are expressed as close. Deviation 14 is defined as the stated difference.

In cases where a given set of reference scenes is associated with a given set of used features 74, a parameterizable transformation 12 (and its parameters) can be selected as the transformed image 13 of each reference scene 9 and the current consideration A function of the difference between the synthetic image classes 7 of each reference scene 9 . There is a selected parameterizable transformation 12 (and its parameters) that can be used to transform the reference image 11 into the transformed image 13 such that the sum of said differences is minimized. The sum function can be replaced by another function, such as multiply. Then, the synthetic image class 7 and the transformed image 13 are expressed as close. The deviation 14 is defined as the value obtained from said difference, for example by calculating its mean value.

In some cases where the used features 74 are unknown, it is possible to determine these features from the capture of a plurality of reference images 11 of at least one reference scene 9 . In this case, the unknown feature and the parameterizable transformation 12 (and its parameters) are determined simultaneously so that they can be used to transform the reference image 11 into the transformed image 13 such that the sum of said differences is minimized, specifically is calculated iteratively, or solving equations for the sum and/or product of said differences and/or any other suitable combination of said differences. Then, the synthetic image class 7 and the transformed image 13 are expressed as close. For example, the unknown feature may be the relative position and orientation of the sensor surface 110 and the currently considered reference scene 9 . The deviation 14 is defined as the value obtained from said difference, for example by calculating its mean value.

Best Transformation

An optimal transformation is defined as a transformation in a parameterizable transformation model 12 by which each reference image 11 can be transformed into a composite image class 7 corresponding to the reference scene 9 corresponding to said reference image 11. Close transform image13.

Calibration

Calibration is defined as a method by which data relating to the intrinsic characteristics of the image capture device 1 can be obtained for one or more of the configurations used, where each configuration contains a configuration associated with the image capture device 1 Optical system 100.

Case 1: This case contains only one configuration, and the method includes the following steps:

The step of installing the optical system 100 on the image capture device 1,

Steps to select one or more reference scenarios 9,

Steps to select several used features 74,

the step of capturing an image of said reference scene 9 for said used feature,

A step of calculating an optimal transformation for each set of reference scenes 9 corresponding to the same used features 74 .

Case 2: This case considers all configurations corresponding to a given image capture device 1 and all optical systems 100 of the same type, the method comprising the following steps:

the step of selecting one or more reference scenarios 9;

Steps for selecting several features 74 for use;

The step of calculating the image 103 from the features 74 used, in particular from the scheme of the optical system 100 of the configuration used, and from the values of the parameters, by means of software for calculating the optical system, for example by ray tracing,

A step of calculating an optimal transformation for each set of reference scenes 9 corresponding to the same used features.

Case 3: This case considers all configurations corresponding to a given optical system 100 and all image capture devices 1 of the same type, said method comprising the following steps:

The step of mounting said optical system 100 on an image capture device 1 of the type currently under consideration,

Steps to select one or more reference scenarios 9,

Steps to select several used features 74,

the step of acquiring an image of said reference scene 9 for said used feature,

A step of calculating an optimal transformation for each set of reference scenes 9 corresponding to the same used features.

Preferably, each device and configuration in case 1 can be calibrated by the manufacturer of the image capture device 1 . This approach is more accurate, but adds more constraints, and is highly applicable when the optical systems 100 are not interchangeable.

Alternatively, calibration for each device type and configuration in Case 2 may be performed by the manufacturer of the image capture device 1 . This method is less precise, but simpler.

Alternatively, each device type and optical system 100 in case 3 may be calibrated by the manufacturer of the image capture device 1 . This approach is a compromise where one optical system 100 can be used on all image capture devices 1 of one type without repeating the calibration for each image capture device 1 and optical system 100 combination.

Alternatively, calibration may be performed by the device vendor or installer for each image capture device and configuration in Case 1.

Alternatively, for each optical system 100 and each type of device in Case 3, calibration may be performed by the seller or installer of the device.

Alternatively, calibration may be performed by the user of the device for each device and configuration in case 1 .

Alternatively, for each optical system 100 and each type of device in case 3, calibration may be performed by the user of the device.

Design of digital optical system

Design of a digital optical system is defined as the method used to reduce the cost of the optical system 100 by:

Designing an optical system 100 with defects, in particular true point positioning defects, or selecting the above systems from the catalogue,

reduce the number of lenses, and/or

simplify the shape of the lens, and/or

Use less expensive materials, handling operations or manufacturing processes.

The method comprises the steps of:

the step of selecting an acceptable variance (within the meaning defined above),

Steps to select one or more reference scenarios 9,

Steps to select several used features 74,

The method also includes repetitions of the steps of:

the step of selecting an optical solution which specifically includes the shape, material and arrangement of the lenses,

By using, for example, software for calculating the optical system by ray tracing, or by taking measurements on a prototype, the step of calculating the image 103 from the features 74 used, in particular from the scheme of the optical system 100 configured in use,

a step of calculating the best transformation for each set of reference scenes 9 corresponding to the same used features 74,

Steps to verify that the difference is acceptable until the difference is acceptable.

Formatting information

Formatting information 15 associated with image 103, or formatting information 15, is defined as all or part of the following data:

data relating to inherent technical characteristics of the image capture device 1, in particular distortion characteristics, and/or

data relating to the technical characteristics of the image capture device 1 at the time of capturing the image, in particular the exposure time, and/or

data relating to said user preferences, in particular color temperature, and/or

Data involving deviation 14.

Feature database

A feature database 22 or database 22 is defined as a database for one or more image capture devices 1 and one or more images 103 containing formatted information 15 .

The feature database 22 can be stored centrally or distributed, in particular it can:

integrated into the image capture device 1,

integrated into the optical system 100,

integrated into removable storage devices,

integrated into a PC or other computer connected to other components during image capture,

After the image capture is completed, it is integrated into a PC or other computer connected to other components,

integrated into a PC or other computer capable of reading the storage medium shared with the image capture device 1,

Integrated into a remote server connected to a PC or other computer, itself connected to other image capture elements.

field

Referring to Figure 8, the definition of field 91 will now be described. The formatting information 15 associated with the image 103 can be recorded in several forms and structured into one or more tables. But it logically corresponds to all or parts of field 90, including:

(a) focal length,

(b) depth of field,

(c) Geometric defects.

Said geometrical imperfections include geometrical imperfections of the image 103, characterized by parameters associated with the captured features 74 and a parameterizable transformation representing the characteristics of the image capture device 1 at the time of capturing. By means of said parameters and said parametric transformation it is possible to calculate the corrected position of a point of image 103 .

The geometrical defects also include vignetting, characterized by parameters associated with the captured feature 74 and a parameterizable transformation representing the characteristics of the image capture device 1 at the time of the captured image. By means of said parameters and said parametric transformation it is possible to calculate the correction strength of a point of the image 103 .

Said geometrical defects also include color casts, characterized by parameters associated with the captured feature 74 and a parametric transformation representative of the characteristics of the image capture device 1 at the time of capture. By means of said parameters and said parametric transformation it is possible to calculate the corrected color of a point of the image 103 .

The field 90 also includes (d) the sharpness of the image 103 .

The sharpness, including blurring in the resolution of the image 103, is characterized by a parameter associated with the capture feature 74 and a parameterizable transformation representing the characteristics of the image capture device 1 at the time of capture. By means of said parameters and said parametric transformation it is possible to calculate the corrected shape of a point of the image 103 . Blurring specifically includes: coma, spherical aberration, astigmatism, grouping into pixels 104, chromatic aberration, depth of field, diffraction, parasitic reflection, and field curvature.

The sharpness also includes blurring in depth of field, especially spherical aberration, coma, astigmatism. The ambiguity depends on the distance of points in the scene 3 relative to the image capture device 1 , characterized by parameters associated with the captured features 74 and a parameterizable transformation representing the characteristics of the image capture device 1 at the time of capture. By means of said parameters and said parametric transformation it is possible to calculate the corrected shape of a point of the image 103 .

Said field 90 also includes (e) parameters of the quantization method. Said parameters depend on the geometry and physics of the sensor 101, the architecture of the electronics unit 102 and any processing software that may be used.

Said parameters comprise variations representing the intensity of pixels 104 as a function of wavelength and luminous flux obtained from said scene 3 . The function specifically includes gamma information.

The parameters also include:

The geometry of the sensor 101, especially the shape, relative position and number of sensing elements of the sensor 101,

A function representing the spatial and temporal distribution of the noise of the image capture device 1,

A value representing the exposure time used for image capture.

Said field 90 also includes (f) digital processing operations performed by the image capture device 1, in particular digital zooming and compression. These parameters depend on the processing software of the image capture device 1 and on user adjustments.

The field 90 also includes:

The (g) parameter represents the user's preferences, especially with regard to the degree of blur and resolution of the image 103.

(h) Deviation 14.

Calculation of formatting information

Formatting information 15 can be calculated and stored in database 22 in several steps.

a) Steps at the end of the design of the image capture device 1 .

By means of this step, it is possible to obtain the inherent technical features of the image capture device 1, in particular:

the spatial and temporal distribution of the noise generated by the electronic unit 102,

Conversion formula from luminous flux to pixel value.

Geometry of sensor 101 .

b) Steps at the end of the calibration or design of the digital optical system.

By means of this step, it is possible to obtain other inherent technical features of the image capture device 1, in particular an optimal correlation transformation and correlation bias 14 for a certain number of used eigenvalues.

c) Step of selecting the user's preferences by means of buttons, menus or removable media or by means of a connection to another device.

d) Image capture step

By means of this step (d), it is possible to obtain the technical characteristics of the image capture device 1 at the time of capturing the image, in particular the exposure time, which is determined manually or by the automatic adjustment performed.

By means of step (d) it is also possible to obtain the focal length. The focal length is calculated by

a measurement of the position of the set of variable focal length lenses of optical system 100 in the configuration used, or

Enter a setpoint for the positioning motor, or

The value given by the manufacturer, if the focal length is fixed

The focal distance can then be determined by analyzing the content of the image 103 .

By means of step (d) it is also possible to obtain the depth of field. Depth of field is calculated by

a measurement of the position of the set of focusing lenses of optical system 100 in the configuration used, or

Enter a setpoint for the positioning motor, or

The value given by the manufacturer if the depth of field is fixed.

By means of step (d) it is also possible to obtain geometric and sharpness defects. The geometric and sharpness defects correspond to a transformation calculated by means of a combination of the transformations of the feature database 22 obtained at the end of stage (b). This combination was chosen to represent the parameter values corresponding to the features 74 used, especially the focal length.

By means of step (d), it is also possible to obtain parameters of the digital processing performed by the image capture device 1 . These parameters are determined manually or through automatic adjustments performed.

According to stages (a) to (d), the calculation of the formatting information 15 can be done by:

equipment or software integrated into the image capture device 1, and/or

driver software in a PC or other computer, and/or

software on a PC or other computer, and/or

A combination of the above three items.

The transformations in the preceding stages (b) and (d) can be stored as follows:

general math formulas,

The mathematical formula corresponding to each point,

Mathematical formulas corresponding to certain feature points,

This mathematical formula can be described by:

list of coefficients;

list of coefficients and coordinates;

With these different methods, a compromise can be reached between the memory size available for storing the formulas and the computing power available for calculating the corrected image 71 .

In addition, in order to retrieve data, identifiers associated with these data are also recorded in the database 22 . Specifically, these identifiers include:

an identifier of the type and index of the image capture device 1,

If the optical system is removable, an identifier of the type and index of the optical system 100,

An identifier of the type and index of any other removable element that is linked to the stored information.

the identifier of the image 103,

Identifier for formatted message 15.

Completed image

As depicted in FIG. 11 , the finished image 120 is defined as the image 103 associated with the formatting information 15 . Preferably, the completed image 120 may be in the form of a file, as shown in FIG. 14 . The finished image 120 can also be divided into multiple files.

The image 120 can be computed by the image capture device 1 . The image can also be computed by an external computing device, such as a computer.

Image processing software

Image processing software 4 is defined as software that accepts one or more finished images 120 as input and performs processing operations on these images. Specifically, these processing operations may include:

Computing a correction image 71,

measure in the real world,

Combine several images,

improve the fidelity of images relative to the real world,

improve the subjective quality of the image,

detecting an object or person 107 in scene 3,

Add object or person 107 in scene 3,

replacing or modifying an object or person 107 in scene 3,

Remove shadows from scene 3,

Add shadows to scene 3,

Search for objects in the image library.

The image processing software can:

integrated into the image capture device 1,

It runs on a computing device 17 connected to the image capture device 1 via a transmission device 18 .

Digital Optical System

A digital optical system is defined as the combination of image capture device 1, feature database 22, and computing device 17 that allows:

image capture of image 103,

The calculation completes the image,

Computing the corrected image 71,

Preferably, the user obtains the corrected image 71 directly. The user can ask to suppress the automatic correction if desired.

The feature database 22 can:

integrated into the image capture device 1,

integrated into a PC or other computer connected to other components during image capture,

Integration into a PC or other computer connected to other components after image capture,

integrated into a PC or other computer that can read the storage medium shared with the image capture device 1,

Integrated into a remote server connected to a PC or other computer, the feature database 22 itself is connected to other image capture elements.

Computing device 17 may:

integrated into a component together with the sensor 101,

together with the parts of the electronic unit 102 are integrated into an assembly,

integrated into the image capture device 1,

integrated into a PC or other computer connected to other components during image capture,

Integration into a PC or other computer connected to other components after image capture,

integrated into a PC or other computer capable of reading the storage medium shared with the image capture device 1,

Integrated into a remote server connected to a PC or other computer, the computing device 17 itself is connected to other image capture elements.

Processing of complete chains

The above paragraphs basically give precise details of the concept and description of the method and system according to the invention for providing formatted information 15 relating to features of the image capture device 1 to the image processing software 4 .

In the following paragraphs, an extended definition of the concept is given, as well as a supplementary description of the method and system for providing image processing software 4 with formatted information 15 relating to the features of the image restoration device 19 according to the invention. The processing of the complete chain will be explained in this way.

With the help of processing the complete chain it is possible to:

improving the quality of the image 103 from one end of the chain to the other, obtaining a restored image 191 while correcting defects in the image capture device 1 and image restoration device 19, and/or

In video projectors, lower quality and lower cost optical systems are used in conjunction with software to improve image quality.

Restrictions associated with image restoration devices

Based on FIGS. 2 and 6, it will now be described how in formatting information 15 the characteristics of an image restoration device 19, such as a printer, a visual display screen or a projector, are taken into account.

Using the analogy provided in the case of the image capture device 1 , a person skilled in the art can deduce by analogy how to supplement and modify this restriction in the case of the image restoration device 19 . However, in order to clarify the method, with specific reference now to Figure 6, the main additions or modifications are described.

The restoration feature 95 used specifies the inherent characteristics of the image restoration device 19, the characteristics of the image restoration device 19 at the time of image restoration, and the user's preferences at the time of image restoration. Particularly in the case of projectors, the restoration features 95 used include the shape and position of the screen used.

A mathematical transformation similar to the parameterizable transformation model 12 is specified by the parameterizable restoration transformation model 97 (or simply called the parametric restoration transformation 97).

The rectified restored image 94 specifies the image obtained by applying a parametric restored transformation 97 to the image 103 .

Mathematical restoration projection 96 specifies a mathematical projection that relates corrected restoration image 94 to mathematical restoration image 92 on a mathematical restoration surface geometrically associated with the surface of restoration medium 190 . The mathematically restored points of the mathematically restored surface have shapes, positions, colors and intensities calculated from the corrected restored image 94 .

Real restoration projection 90 specifies the projection that links restoration image 191 to image 103 . The pixel values of the image 103 are converted by the electronic unit of the restoration means 19 into signals driving the modulators of the restoration means 19 . A real recovery point is obtained on the recovery medium 190 . The characteristics of the true recovery point include shape, color, intensity and location. The above-described phenomenon of grouping into pixels 104 in the case of the image capture device 1 does not occur in the case of the image restoration device. On the other hand, the opposite phenomenon occurs, with the specific result that straight lines appear stepped.

Restoration difference 93 is specified as the difference between restoration image 191 and mathematical restoration image 92 . This recovery difference 93 is derived by analogy from the difference 73 .

The restoration reference specifies an image 103 in which the values of pixels 104 are known.

The optimal restoration transformation specifies the restoration reference and the restoration features 95 used with which the image 103 can be transformed into a corrected restoration image 94 such that its mathematical restoration projection 92 exhibits the smallest restoration difference 93 compared to the restoration image 191 .

In the case of the image capture device 1, the recovery calibration method and design method of the digital optical recovery system can be compared with the correction method and the design method of the digital optical system. However, there are differences in some steps, especially in the following steps:

Select steps to restore references;

Steps for performing recovery of said recovery reference;

Steps to compute the optimal restoration transform.

The formatting information 15 relating to the image capture device 1 as well as to the image restoration device 19 can be used end to end for the same image.

The concept of a field in the case of the image capture device 1 has been given above. This concept also applies by analogy to the case of the image restoration device 19 . However, the parameters of the quantization method are replaced by parameters of the signal reconstruction method, namely: the geometry of the restoration medium 190 and its position, a function of the temporal and spatial distribution of the noise of the image restoration device 19 .

Generalization of the concept

The technical features contained in the invention and specified in the claims have been defined, described and exemplified with reference to an image capture device of the digital type, ie a device for generating digital images. It will be readily understood that the same technical features apply in the case of image capture devices combined for devices based on silver technology using scanners that produce digital images from developed photographic film (using sensitive silver halide film, negative or negative photographic or cinematic device) of film. Of course, in such cases, at least some of the definitions used may be appropriately modified. These modifications are within the purview of those skilled in the art. To demonstrate the obviousness of this modification, it is only necessary to mention that in the case of a combination of devices based on silver technology with a scanner, after digitization of the unit area of the film surface by means of a scanner, with reference to the image shown in FIG. 3 The concepts of pixels and pixel values must be applied to the cell area. This defined inversion is self-evident and can be extended to the concept of the configuration used. For example, the list of removable components of the image capture device 1 included in the configuration used may also be supplemented by the types of photographic film that are effectively used in devices based on silver technology.

Automatic modification of image identity

Referring to Figures 17, 13a, 13b and 13c, an example of a method and system according to the present invention for the automatic improvement of the quality of an image P2 originating from or destined for a chain of devices P3, which is also defined as specifying device chain.

In this example, the system according to the invention comprises data processing means P76 for performing the following steps of the method according to the invention on the image P2:

Step of editing the directory of sources P50 of formatted information 15 related to devices of device group P75; in particular, these sources P50 may be image files P58 containing images P2, devices, local and/or remote databases 22, depending on the situation , and the means P53 for loading images P2 or modified images, an example being dual adaptation software for scanners; sources edited in this way are limited to databases 22,

Among the formatted information compiled in this way, the specific formatted information related to the specified device chain P3 is automatically retrieved; a source can be used to update the database 22, for example via the Internet,

In case the retrieval of specific formatting information P51 for one of the devices is unsuccessful, for example by measuring a defect of the device, or by simulating the device, or by calibrating the device according to the method described above, a step of calculating the unknown formatting information,

The step of automatically modifying said image P2 by means of image processing software P64 and/or image processing components P65 by taking into account said specific formatting information P51 obtained in this way.

By way of example, the application of the method and system according to the invention to the restoration of images of an image capture device, whether it be a photographic device, a video capture device, an echographic device, etc., will be described. In these cases it will be possible to use photographing devices (especially of the optical type), sensors or photosensitive surfaces (CCD, photosensitive film, etc.), scanners, processing software, information transmission chains between different characteristics of the printer. As will be seen below, it will be possible to use other elements in the image chain with their own specific characteristics.

Each device is characterized by an identifier 60, by means of which it is possible to identify this type of device and thus access the known characteristics associated with this type of device and indirectly obtain the index P52, the application of which will be described below .

Certain units may be subject to special operating adjustments. For example, the capture device can be adjusted according to the shooting conditions, or the printer can be set to enlarge mode, etc. These specific modes of operation lead to variable feature P55, which can be used within the scope of the present invention. Furthermore, these variable characteristics P55 may affect the fixed characteristics CS (or original characteristics) of the device or device chain P3.

Figure 13a shows a schematic diagram of a practical example of the system of the present invention.

In the upper part of the diagram, devices or devices are shown, defined as peripheral devices, which provide information, images and data to the system:

Involved are:

- the image 1 to be processed, which obviously includes image information 61, but may also contain data 62, which relate to the characteristics of the medium in which it was captured and which provide information about the conditions of image capture or information about the subsequent processing of the image ( such as the focal length used, or re-scaled images). These features will be referred to as variable features 66 in the following description. The variable feature can also be contained in the image itself or can be calculated from the image.

- The image capture device APP1 , which has inherent features and which may contain variable features 66 . The inherent characteristics are related to the type of device or to each device, and these characteristics can be known through knowledge of the device and knowledge of the original characteristics. In particular, they are known by their identifier 60, which may be, for example, a barcode on the device or film, and which are used by the system to associate these inherent characteristics. The identifier 60 can be obtained from the image in a number of ways, from the data 62, and/or by interrogating the device's management software or the device itself, or the user, these are represented in Figure 13a by the LOG/MAT/IMPR notation change. Variable features 66 are generally related to image capture conditions, and may be those described above, which may be included in the image, or in the image capture device, or both:

- Peripherals APP2 to APPn, including in particular peripherals APP2 such as printers or scanners, which peripherals APP2 have characteristics related to the type of devices and in particular reflecting their deficiencies, and variable characteristics related to the mode of use 66, such as a magnification factor for a printer. Peripherals such as APPn may also be pseudo-peripherals and be in the form of files representing these means or functions, and containing features corresponding to these means or functions:

image capture device,

image restoration device,

an image capture device other than device APP1, used to simulate a picture taken by the device, and which may be called "lookalike" in English,

The simulation of the image capture device, which is different from the device APP1, is used to replace the above-mentioned device, and by analogy from the previous situation, it can be called a virtual "lookalike",

Features used to improve image quality, or to generate special effects, by taking into account visual imperfections of the user's naked eye.

Software defects such as blurring caused by digital zoom can be corrected or modified features,

Software that manipulates images and introduces defects, or is provided with variable features66, such as recording or zooming,

Mounting characteristics, especially for projectors, which are related to errors in square on frontal delineation or errors in screen flatness, and which can be measured by a camera, for example,

A combination of multiple devices as described above.

These various devices APP1 to APPn and image 1 comprise a device chain P3. The processing of the image will take into account not only its inherent imperfections, but also the imperfections of different devices, all of which will act on the image. These devices of this group will be defined as device chain P3. The purpose of the system is to correct or modify defects introduced into the image processing by each device.

Examples of variable features 66 that may be considered appropriate include:

the focal length of the optical system,

Resizing applied to the image (digital magnification: enlargement of part of the image; and/or undersampling: reduction of the number of pixels in the image),

non-linear brightness corrections, such as gamma correction,

Contour enhancements, such as the level of deblurring applied to the device,

the noise of the sensor and the noise of the electronic unit,

the aperture of the optical system,

focal length,

the number of frames on the film,

underexposed or overexposed,

film or sensor sensitivity,

the type of paper used in the printer,

The center position of the sensor in the image,

The rotation of the image relative to the sensor,

the position of the projector relative to the screen,

The white balance used,

flashes and/or their energies of power,

exposure time,

sensor gain,

compression,

Compared,

another adjustment imposed by the user of the device, such as the mode of operation,

Another automatic adjustment to the device,

Another measurement implemented by the device.

The system has reception interfaces C.VAR1 , C.VAR1 , . . . C.VARn, which are designed to receive the variable characteristics 66 explained above. Features derived from the image may be conveyed through the image itself, or, as described above, through data associated with the image. Recall that the variable characteristics 66 of the image can also be transmitted by the image capture means.

The interfaces ID1, ID2, . . . IDn are designed to receive identifiers 60 of the different peripheral devices APP1-APPn.

Depending on the case, the concept of a peripheral may correspond to one or more devices, which may or may not be of the same type. The following examples, each with possible implementations of identifiers 60 in the form of codes and contents, correspond to several of these cases:

For a given peripheral device (for example, code IA1: manufacturer's name, peripheral device type, serial number of peripheral device),

For a given type of peripheral (eg code IA2: manufacturer name, peripheral type)

For a given configuration (eg code IA3: manufacturer name, peripheral type, type of interchangeable objects installed)

For classification of peripherals (e.g. code IA4, for one-time photographic devices: name of manufacturer, type of peripheral, number of frames),

For manufacturers (eg code IA5: for manufacturers)

For a chain of multiple peripherals (for example, code IA6 in the item of equipment for photo IC printing, import formatted information items related to disposables and combine these information items with scanners and printers for stored in a local database associated with the chain, requiring the applicable identifier 60),

For algorithms, such as scaling algorithms (eg, code IA7: Algorithm Name, Implementation, field indicating whether the image quality was modified before or after this algorithm),

For the person whose visual impairment is to be corrected or amended (e.g. code IA8, person's name, country),

For peripherals to be emulated and peripherals to which adding defects does not suppress defects (for example, code IA9: manufacturer name, device type),

For a given peripheral version (eg code IA10: manufacturer name, peripheral type, peripheral software version)

For protocols (for example, encoding IA11: data obtained from the Twain protocol),

For general peripherals (eg code IA12: list of data sources, field identifier; field value).

The system can then analyze the peripherals or devices of the device chain P3 and, depending on the case, determine the identifier 60 differently in order to be able to query the database.

For each type of device, the database includes at least one item of formatted information representative of defects and characteristics of the device. The formatting information 15 may be related to the defect P5 of the device in several ways. It may represent a defect of the device. It may represent the opposite of the flaw of the device. It may only represent an approximation of this defect. It may represent a deviation between the defects of the two devices. From each identifier 60 provided by an interface, for example interface ID1, it is possible to obtain an item of formatted information, for example 15.1, temporarily received in circuit 20.1. Formatted information related to devices APP1, APP2, APP3... APPn may be received in circuits 20.1, 20.2, ... 20.n.

The database may be integrated into the system or may be at least partially remote.

In this case, at least in part, it may be managed by a third party.

In the case, for example, of a digital camera correcting or modifying its inherent flaws, the database can be reduced to a registry.

Operators 21.1, 21.2, . The formatted information is provided in such a way that each operator processes a part of the formatted information by means of one or more variable features 66 and provides modified formatted information items to the intermediate circuits 22.1 to 22.n. For example, the operator 21.1 receives the formatted information item 15.1 according to the focal length, processes it by means of the value of the variable feature 66 (focal length value) provided by the interface C.VAR1 and provides a modified format independent of the focal length Information item 15.1'. For example, Figure 16.1 shows that the device possesses features that make a defect 704 apparent, which subsequently results in formatting information 15, as seen above. A variable feature 66, representing a variable focal length 705, for example also results in formatting information (see Figure 16.2).

Image processing operators 23.1, 23.2, ... 23.n are each designed to receive modified formatted information items. The first operator 23.1 receives the image to be processed, processes the image by means of the modified formatted information item 15.1' and provides the modified image. This is received by the following operator 23.2, which processes the image by means of the modified formatted information item 15.2' and provides a new modified image, etc., until the last operator 23.n , which provides the final modified image.

Possibly, if an image processing operator does not receive any modified formatting information, the image received by such operator is not modified and transmitted to subsequent operators or output, or for example default formatting information can be used .

Finally, the entire operation of the system, in particular the exchange of information and data between the different elements of the system, can be managed by a central control unit 25 .

Under these conditions, the central control unit 25 will manage the automatic retrieval in the database 22 of formatted information having addresses given by the interfaces ID1, ID2, . . . Idn. The central unit 25 manages the retrieval of this information and activates operators 21.1 to 21.n for processing the formatted information and then activates image processing operators 23.1 to 23.n. Operators can be located in different remote but interconnected systems if desired.

Figure 13b shows an alternative embodiment of the system according to the invention. In this alternative embodiment, the modified formatting information items are combined into a single formatting information item and they improve the quality of the image to be processed. For this purpose, one operator 23.t replaces the operators 23.1 to 23.n. The operator receives different modified items of formatted information and combines them and makes it possible to improve the quality of the image to be processed in order to provide a modified image.

Furthermore, according to an alternative embodiment also applied to the system of Fig. 13a, as shown in Fig. 13c, wherein the variable characteristics 66 of the device and its identifier are combined to enable direct access within the database 22 to the modified formatted information item. For example, the variable feature 66 provided by C.VAR1 and the identifier ID1 are combined to form a modified formatted information item 15.1, which is transmitted to 22.1. Clearly, in Fig. 13c, this setting is provided only for the modified formatting information item 15.1', but it may be applied to all or part of the other formatting information. The system provides modified images at the output of the operator 23.n for Figures 13a and 23, and Figures 13b and 13c. Figure 16.5 shows this situation, where a modified image 61c is added with relevant information 62c, which may be:

digital signature of the image,

correction data, or in other words, modified formatting information or its equivalent, or simply an indicator whereby the image has been corrected or modified,

Data 62 or information P63 associated with the original image 1, modified or updated if necessary to reflect the processing operations applied to the image, e.g. data in Exif or PIM format,

or both types of data.

The practical examples of Figures 13a to 13c can be used for all or every defect.

For example, it is possible to correct the distortion and blurring of the device APP1, then the distortion and blurring of APP2, and so on.

According to another example, it is possible to correct the distortion of APP1, then the distortion of APP2, etc., then the blurring of APP1, then the blurring of APP2, etc.

By generalizing in the case of multiple defects and multiple devices, it is possible to combine the two

Example method.

Referring to Figure 14a, a simplified example of the method of the present invention will be described. This example applies to image capture devices. It is assumed that the method is required to correct only defects due to a single device, such as an image capture device, and defects due to adjustments to that device.

As an example, the image may be processed in digitized form by digitizing device 400.1, by digital capture device 400.2 (digital camera, or scanner or other device), or by compact disc 400.3.

In step 401 of the method, a digitized image may be obtained. Furthermore, the identity of the image capture device and even the type of the device can be known by any means of identification, such as a barcode.

In step 402 of the method, an identifier 60 of the device is obtained or calculated.

In step 403, the feature database 22 of the image capture device can be accessed via the identifier 60, for example by means of the index P52. In fact, as mentioned above, a database is available in which, in principle, device characteristics have been registered for each known device. These features represent drawbacks to be corrected within the scope of the invention. The database 22 is thus called up in step 403 . In an alternative embodiment, the call to the database may additionally take into account some of the variable features 66 obtained at step 405 to directly obtain formatting information that, for variable features 66, is useful for The values obtained in this way are relative.

In step 404 , an item of formatted information 15 representative of the characteristics (defects) of the corresponding device is read in the database 22 located at the address obtained from the identifier 60 .

Furthermore, together with the image to be processed, if necessary, the system can be provided with variable characteristics 66 from the memory of the device, from software associated with the device, or from information related to the image (in particular the photographic condition).

These features are thus obtained in step 405 .

Thereafter (step 406), the formatted information 15.1 to 15.n is combined with variable features 66 to provide modified formats 15.1' to 15.n'. This modified formatting information now contains all information that can be used to improve image quality.

According to an alternative version of this step 406, if the variable characteristics 66 and in particular the values of the variable characteristics 66 for the image to be processed have been determined, they are used to determine the consideration of these variable characteristics within the formatted information 66 part of this formatting information.

In step 407, this modified formatting information is applied to the image to process it and correct or modify it. This processing is carried out by an operator assisted by image processing software.

In this way a corrected image is obtained in step 408 .

Clearly, by using only the inherent features of the device without using the variable features 66, the foregoing method works. In this case, the image can be processed directly using the formatted data read from the database.

Referring to Fig. 14b, another example of the present invention will be described. In this method and system, it is assumed that there are reasons to consider multiple deficiencies of multiple devices, and even of all devices involved in the image processing process.

As described in the example of Fig. 14a, the method provides that the identifier 60 and variable characteristics 66 of the device are captured by the system of digitized images.

During a step 501 , the identifier 60 of the device is taken into account and makes it possible to address the database 22 (step 502 ) in order to obtain one or more items of formatted information corresponding to this identifier 60 .

A search is also made for variable characteristics 66 associated with the device (step 504).

During a step 505 , the formatting information or some items of formatting information are modified according to the characteristic variable 66 . As described in connection with the method of FIG. 14b, once variable features 66 are determined, they can be used to determine which are useful and which take variable features 66 into account in formatting information. Formatting information determined in this way is stored in memory.

Next (step 506), a test is performed to determine whether another device must be considered for improvement in image quality. In the schematic diagram of Figure 14b, this test is represented in the form of the question "APP=APPn?", meaning "Is the device currently under consideration the last in the chain of devices?". If the answer is negative, the process of step 501 is repeated for the next device. If the answer is yes, it means that at the end of step 505, all the formatted information related to the different devices is in memory.

The image is then processed in step 507 by the formatting information associated with the first device, resulting in a first processed image. The system then considers the formatted information associated with the next device and processes the previously processed image, etc., until all the formatted information has been processed, which in principle means until all the different devices of the chain have been considered Related information. The test "APP=APPn?" is then answered in the affirmative. Obtain and deliver the modified image.

It will be noted that the method of the present invention can be implemented by performing only a single test "APP=APPn?". It would be possible to perform only the test of step 508, which would achieve almost the same result.

In an alternative version of the method according to the invention, as shown in FIG. 14c, after obtaining all the formatting information of all devices, or in other words, after step 506 is completed, the provision can be made during step 510 Combine different formatting information items. In this way, image processing is performed at one time during step 507. As noted above, multiple devices may be required for the image to be processed, possibly including image capture devices, scanners, printers, transmission systems, and the like. Each device is capable of introducing defects in the processing chain. Additional means, defined above as "pseudo-peripherals", may be required which modify the image according to or apply the defects corresponding to these pseudo-peripherals.

In an alternative version of the method of the invention, it is considered that in the device chain P3, the set of devices necessary to process the image consists of a single device, which will be defined as a virtual device 708, and whose defects correspond to the different The equivalent of all or part of the defects of the device. Devices such as the image capture device 706 (FIG. 16.4) and the printer 707 can thus be represented by a virtual device 708, which corresponds to the virtual formatting information 709. If it is considered that a formatted information item may be a mathematical representation of a physical characteristic, the formatted information item of the virtual device 708 corresponding to two devices may be a volume corresponding to a feature of the two devices and/or to two mathematical functions The sum of the product of two vectors. According to Fig. 16.4, a virtual device 708 is thus determined which has defects equivalent to at least a part of the chain's original defects. Virtual formatting information corresponding to the virtual device is determined 709 . And register the obtained virtual formatting information, or make the virtual formatting information replace the formatting information related to the original defect. This virtual formatting information is directly accessible in the database by means of the identifier 60 corresponding to the device chain P3 represented by this virtual device. The method can thus be implemented more simply and more quickly.

An example of an organization in which this method can be employed can be obtained from the organization chart of Figure 15, using which virtual formatted information can be obtained. The characteristics of the two devices are considered (steps 510 and 511). These features are combined in step 512 . In step 513 the corresponding virtual formatting information is calculated. In step 514 it is checked whether other devices are required in the virtual device chain. This process is repeated if necessary. If not necessary, the processing is terminated.

In the example of the joint development laboratory, the device chain P3 includes a scanner, a camera and a printer. The equivalent virtual device has the disadvantages of these three devices, and can basically divide the time for improving the image quality into thirds.

In the case of a device having variable features 66, it is possible to determine formatting information related to defects displaying said variable features 66 in the following way.

In the example where the variable feature 66 is focal length and aperture, the following combinations are selected:

focal length = 35mm, aperture = f/2,

focal length = 35mm, aperture = f/8,

focal length = 100mm, aperture = f/8,

etc.

For each combination, the corresponding formatting information is determined by the method described above.

The database contains the necessary formatting information in step 404, for example by inserting formatting information derived as a function of focal length and aperture.

It can thus be seen that, as seen above in the description of Fig. 13a, formatted information items are available which depend on a variable characteristic 66, eg focal length. This item of formatted information is processed by means of variable features 66 to obtain a modified item of formatted information.

In the above description, it is considered that an image to be processed is an image obtained by an image capture device, and is to be displayed or printed. The invention can also be used in any image processing chain and thus also in chains for projecting images. Thus, the image restoration chain will now be considered. As described above in the example of the inventive method, it is necessary to obtain the characteristics of the different means of the image restoration chain. By means of these features it is possible to obtain formatted information for applying the method of the invention.

Details or alternative modifications of the elements of the invention will be described below.

Firstly, during the description of Fig. 13a, it was mentioned that the provided features that make it possible to obtain formatted information may be features designed to correct visual defects of the observer 701 (such as astigmatism), or features that cause special effects. Under these conditions, the formatted information 702 obtained makes it possible to improve the visual, pictorial, chromatic and other qualities of the image; as shown in Figure 16.3, the formatted information 702 related to the visual defect of the observer 701, For example, formatting information handled as the device chain P3 is even related to the formatting information. In the above description, the processing of the image was considered; it is also possible to consider that the image is located in the file P57, together with the identifier 60 or the index P52, and possibly the variable characteristics 66 of the capture device and the image registered in the processing of the file. variable characteristics 66 of any device involved in the image; it is also possible to consider that the image is located in the image file P58 along with part of the formatting information. By extension, the invention is thus equally applicable to the case where images and formatted information are in the database 22 .

The value of this variable feature 66 can be determined by means of information contained in the file P57 or in the image file P58. Preferably, this information will be registered in the file in a standard format, such as the EXIF standard known in the art. In this way, the system and method of the present invention can be used to process images that have been captured and/or have been processed with the aid of a device that was commercialized prior to creating formatted information corresponding to the device.

Obviously, the improvement of the image quality can be simplified by only considering the defects of a limited number of device chains, or even of a single device, and by correcting only these defects.

Furthermore, as already foreseen in the description of Fig. 13a, the method of the invention may be applied by simulating the devices forming part of the device chain P3 being used. Likewise, formatting information associated with one device or type of device may be applied to another device or type of device, especially similar devices. For example, as shown in Figure 16.6, multiple sets of devices 710.0, 710.1, 710.2 are shown. The formatted information relates to a type of device 711, but the formatted information can also be applied to a similar device 712, thereby allowing, for example, only generating formatted information related to each individual device.

The invention is applicable to the modification, or in particular improvement, of images processed or provided by a device or chain of devices P3. One attractive application would be to modify only the defects or partial defects of only certain devices. One application may be to modify defects only partially, for example, to create a trade-off between image quality and computation time.

Another object of the invention is its application to machines for processing images in such a way as to cause defects, to give the image a certain style or to simulate the presentation of a device defined as a reference device, or to simulate a device defined as a reference device Presentation of chains of device chains P3 that differ from the devices and device chains used in the scope of the present application.

The invention can be applied to the design of joint photo development laboratories. It can be employed on a computer.

Finally, the invention can be employed in projectors, in which case different connections are allowed, including parallax correction common in projection technology. For this purpose, a camera or camera can be used to capture the test pattern projected on the screen.

Other examples of device chains may include:

Computer camera (WEBCAM in English),

scanner,

digital camera,

camera,

printer,

Screen,

projector,

game,

image processing software,

teleconferencing system,

Surveillance cameras.

The following examples form a device chain:

single device,

Image capture device and image restoration device,

For example photographic devices, scanners or printers in photo-printing microlabs,

such as digital photographic devices or printers in photo-printing microlabs,

such as a scanner, screen or printer in a computer,

screens or projectors, and the human eye,

expect to imitate one device and another,

photographic devices and scanners,

Image capture device and image processing software,

Image processing software and image restoration device,

A combination of previous examples,

Another set of devices.

The method can be implemented in different forms:

operating system,

Extensions of processing software, such as processing software known under the trademark "PHOTOSHOP",

embedded software,

integrated electronics,

services on the Internet,

Or any combination of these forms of adoption, etc.

Color image

Referring specifically to FIG. 18, the concept of the color image P21, the concept of the color plane P20, the concept of the specified color P22, and the concept of the data P23 related to the specified color will be described below. An alternative embodiment described below is applicable to the case where the image P2 is a color image P21. Image P21 can be decomposed into color planes P20 in a number of ways: number of said planes (1, 3 or more), precision (8-bit unsigned, 16-bit signed, floating point, etc.), and all The importance of the described plane (relative to the standard color space). Image P21 can be decomposed into color planes P20 by various methods: red, green, blue (RGB), or brightness, saturation, hue, etc.; on the other hand, there are color spaces such as PIM, or there may be negative images Pixel values in order to be able to represent subtractive colors, which cannot be represented in positive RGB; finally, it is possible to encode pixel values using 8-bit, 16-bit or floating-point values. The formatting information 15 comprises data with which the image P2 can be decomposed into color planes P20 compatible with the different defects P5 to be processed; each color plane is characterized by an assigned color P22; said formatting information 15 contains data P23 related to said specified color, such as coordinates in the standard CIE or XYZ or LAB or sRGB color space, which data P23 related to said specified color make it possible to calculate the color plane P20 of the image 1, And determine a part of the formatting information 15, this part of the visualization information can be used to properly improve the quality of the color plane P20.

In the case of devices compatible with the PIM standard, it is possible, for example, to choose to operate in 8-bit orthochromatic in X, Y, Z space, or in signed 16-bit in RGB space.

  Standard formatting information, extended formatting information

As shown in FIG. 15 , formatted information 15 or portions of formatted information 15 may contain standard formatted information P101 to describe raw measurements, such as mathematical fields related to geometric deformation defects at a certain number of feature points of array 80 . The formatted information 15 or part of the formatted information 15 may include extended formatted information P102 by, for example, inserting real points of the array 80 instead of feature points. calculated. From the above it can be seen that the formatted information item 15 can rely on the variable feature 66 . According to the invention, a combination P120 is defined as a combination of variable features 66 and their values, such as a combination P120 of focal length, focusing, aperture aperture, capture speed, aperture, etc., and related values. It is difficult to imagine how to calculate the formatting information 15 involving different combinations P120, especially since some characteristics of the combinations P120, such as focal length and distance, may vary continuously.

The invention provides for calculating formatting information 15 in the form of extended formatting information P102 by interpolation from standard formatting information P101 relating to a predetermined selection of combinations P120 of known variable features 66 .

For example, the standard formatting information P101 related to the following combination P120 is used to calculate the extended formatting information P102 dependent on the focal length as the variable feature 66: a combination including "focal length = 2, distance = 7, capture speed = 1/100" , including the combination of "focal length = 10, distance = 7, capture speed = 1/100", including the combination of "focal length = 50, distance = 7, capture speed = 1/100". Through this extended formatting information P102, it is specifically possible to determine formatting information related to the following combination: "focal length = 25, distance = 7, capture speed = 1/100".

There may be an insertion deviation P121 between the standard formatting information P101 and the extended formatting information P102. The invention may comprise the step of selecting 0, 1 or more variable features 66 such that the insertion deviation P121 for the obtained extended formatting information P102 for the variable features 66 selected in this way is less than a predetermined insertion threshold . In fact, some variable features 66 have less influence on defect P5 than others, and the error introduced by treating these variable features 66 approximately as constants may be minimal. For example, focus adjustments have only a slight effect on vignetting artifacts and may not be part of the selected variable feature 66 for this reason. Variable features 66 may be selected at the time formatted information 15 is generated. The combination of technical features enables the improvement of image quality with simple calculations. The combination of technical features also makes the extended formatting information P102 compact. The combination of technical features also minimizes the impact of the eliminated variable features 66 on the defect P5. The combination of technical features also makes it possible to improve the image quality by means of the formatted information 15 with specified precision.

Apply the present invention to reduce cost

Cost reduction is defined as a method and system for reducing the cost of a device or a device chain P3, in particular the cost of an optical system of a device or a device chain, the method comprising:

- reduce the number of lenses, and/or

- simplify the shape of the lens, and/or

- design an optical system with a defect P5 greater than that expected for the device or device chain, or select the same optical system from the catalog, and/or

- Adoption of materials, components, processing operations or manufacturing methods that reduce the cost of the device or device chain and increase the defect P5.

The method and system according to the invention can be used to reduce the cost of a device or a device chain: a digital optical system can be designed to generate formatted information 15 related to the defect P5 of the device or device chain, to use this formatted information to realize the image Processing means, whether these means are integrated or not, to improve the quality of images originating from or sent to the device or chain of devices, so that a combination of devices or chains of devices with image processing means can capture, at reduced cost, Modify or restore images of desired quality.

Claims (43)

1. A method for improving the quality of at least one image (P2) originating from or being sent to a specified device chain (P3); said device chain (P3) comprising at least one image capture device (1 ) and/or at least one image restoration device (19); said image capture device (1) and/or said image restoration device (19) is composed of different economic participations belonging to a group (P75) of indeterminate devices or gradually commercialized; devices in said group of devices (P75) exhibit defects (P5) that can be characterized by formatted information (15); Said method comprises, for said image (P2), carrying out the following steps: - a step of editing the directory of sources (P50) of formatting information (15) related to said devices of said group of devices (P75), - a step of automatically retrieving, among said formatting information (15) compiled in this way, specific formatting information related to said specified device chain (P3), - A step of automatically modifying said image (P2) by means of image processing software (P64) and/or image processing components (P65), taking into account said specific formatting information obtained in this way. 2. A method according to claim 1; said method enabling said automatic retrieval by means of an index (P52) obtained directly or indirectly from an analysis of: - said image (P2), and/or - said means of said means chain (P3), and/or - means for loading said image (P2) into said image processing software or components (P64, P65), and/or - means for loading the image modified by the image processing software or component into the image restoration means (19). 3. Method according to claim 2, said devices of said device chain (P3) being identified by an identifier (60), in particular a barcode; said means for retrieving said specific formatted information The analysis includes the step of determining said identifier (60). 4. A method according to any one of claims 1-3, said method causing said image (P2), said index (P52) and/or said identifier (60) to be contained in the same file (P57) middle; This makes it possible to empirically exploit the method according to the invention in situations where certain means of the chain have been commercialized before the formatted information ( 15 ) related to them has been established. 5. A method according to any one of claims 1-3, said method causing said image (P2) and at least a part of the specific formatting information (15) to be included in the same image file (P58); It is made possible to automatically retrieve said formatting information (15) in said image file (58). 6. A method according to any one of claims 1-3, said method further comprising the step of previously storing at least part of the formatting information (15) in a database (22); said method further comprising updating said database Step of (22). 7. A method according to any one of claims 1-3, said method causing a device in said chain of devices (P3) to have at least one variable feature (66) dependent on the image (P2), in particular is the focal length; a portion of said specific formatted information (15) is related to a defect (P5) of a device having said variable characteristic (66); said method further comprising the steps of: a step of determining a value for said variable feature (66) of said image; a step of determining said portion of said specific formatting information (15) by taking into account the values obtained in this way for said variable feature (66); Such that applying the method to a device with variable features (66) is equivalent to applying the method to a device without any variable features (66). 8. A method according to claim 7, said image being contained in a file (P57); the method makes it possible, for determining the value of said variable characteristic (66), to use, for example, the Exif standard format present in said data in the file (P57), especially data such as focal length; This makes it possible to empirically exploit the method according to the invention in situations where a device with variable characteristics (66) has been commercialized before the formatting information related thereto has been established. 9. Method according to any one of claims 1-3: said method is such that, for improving the quality of at least one image (P2) originating from or destined for a chain of devices (P3): determining (708) a virtual device exhibiting a defect (P5) comparable to at least a portion of said defect (P5), hereinafter referred to as an original defect (P5), of at least one device of said device chain (P3), determining virtual formatting information (709) associated with a defect (P5) of said virtual device (708), for determining specific formatting information associated with the set of devices of said device chain (P3), using said virtual formatting information (709) to replace specific formatting information associated with the original defect (P5), Such that formatting information (15) is then obtained which is more simplified to use and which can be calculated faster and/or using less storage and/or with greater precision to modify said image (P2) The formatting information for the . 10. A method according to any one of claims 1-3, said method being designed to improve the quality of at least one color plane (P20) of a color image (P21); said color plane (P20) being specified by a The specific formatting information (15) further includes data related to the specified color (P22); in order to modify the image, use the data related to the specified color (P22) A color plane (P20) is calculated with said data associated with said image (P2). 11. The method according to claim 1, said method further comprising, in case the process of retrieving said specific formatting information for one of said devices chain (P3) is unsuccessful, calculating said unknown formatted Step of information (15). 12. The method of claim 11, further comprising: By measuring the defect of the device (P5), and/or By simulating the device A step of calculating said unknown formatted information related to a device of said device chain (P3). 13. A method according to claim 11 or 12; for the image capture device (1) of said device chain (P3), said method further comprising the step of calculating said unknown formatting information (15), said calculating By doing the following: constructing a composite image class (7) from a specified mathematical projection (8) of at least one reference scene (9) onto a surface (10), capturing at least one reference image (11) of each reference scene (9) by said image capture device (1), selection among a set of parameterizable transformation models (12) that can be used to transform said reference image (11) into a transformed image that approximates said synthetic image class (7) of said reference scene (9) (13); said transformed image (13) exhibits a deviation (14) compared to said synthetic image class (7); Said unknown formatting information (15) consists at least in part of parameters of said selected parameterizable transformation model (12). 14. The method of claim 13, further comprising: the step of calculating said deviation (14) between said transformed image (13) and said composite image class (7), a step of associating said deviation (14) with said unknown formatting information (15), makes it possible to derive three-dimensional normalized information about said scene, This makes it possible to combine multiple images obtained from multiple image capture devices (1) that have undergone the same formatting process. 15. Method according to claim 13, which causes a device of said device chain (P3) to have at least one variable characteristic (66) dependent on the image (P2), in particular focus and/or aperture ; a part of said specific formatting information (15) is related to a defect (P5) of a device having said variable feature or a plurality of said variable features (66); each variable feature (66) can be associated with a value is combined to form a combination (P120) consisting of the set of said variable features (66) and said value, Said method also comprises the step of determining said part of said unknown formatting information (15), said determining by means of: Select a predetermined combination (P120), For each of said predetermined combinations (P120), employing the process of repeating the steps of the method according to claim 13 or 14, A process of interpolating said unknown formatting information associated with any combination (P120) is employed based on said unknown formatting information obtained at the end of said iterative process. 16. The method according to claim 11 or 12, for the image restoration device (19) of said device chain, said method further comprises the step of: generating a defect (P5) characterizing the image restoration device (19) Data, in particular distortion characteristics, said unknown formatting information (15) at least partially comprises said data characterizing a defect (P5) of the image restoration device (19). 17. A method according to any one of claims 1-3, which causes said specific format associated with a device or devices of said device chain (P3) to be determined in such a way optimize information so that it can be applied to similar devices (712), Such that only a limited amount of formatting information (15) is required for the method to be employed. 18. A method according to any one of claims 1-3, said method causing said image (P2) to include relevant information (P63), especially a digital signal, in such a way that the step so that it retains or modifies said related information (P63). 19. A method according to any one of claims 1-3, said method further comprising the step of associating information with said modified image, in particular information indicating that it has been modified. 20. A method according to any one of claims 1-3, said method being more specifically designed to improve the visual quality of said image for a viewer, in relation to the means of said chain of means (P3) Said formatted information related to a defect (P5) further comprises formatted information related to said observer's visual characteristics, in particular said observer's eye and/or brain dysfunction. 21. Application of the method according to claim 1, the purpose of which is to improve the performance of the image processing software provided by the image processing software by correcting the effect of at least one defect (P5) of the devices of the device chain (P3) (P64) or the quality of the image (P2) processed by said image processing component (P65), This allows the quality of the processed image to be improved, if not perfect, without relying on expensive equipment. 22. The application according to claim 21, the purpose of the application is to make the quality of the image (P2) processed by the image processing software (P64) or the image processing component (P65) comparable to that used The quality of the images produced by a reference device chain (P3) is comparable. 23. The application according to claim 21, which makes it possible, for a processed image of a quality comparable to that produced using a reference device chain, to produce a result similar to that of Formatted information (15) associated with said device chain (P3). 24. A system for improving the quality of at least one image (P2) originating from or being sent to a specific device chain (P3), said device chain (P3) comprising at least one image capture device (1) and/or at least one image restoration device (19), said image capture device (1) and/or image restoration device (19) being gradually commercialized by different economic actors belonging to a group (P75) of indeterminate devices wherein said device of said device group (P75) exhibits a defect (P5) characterizable by formatted information (15), Said system comprises, for said image data processing means (P76), it can: editing a catalog of sources (P50) of formatting information (15) related to said devices of said device group (P75), Among said formatting information (15) edited in this way, specific formatting information related to said designated device chain (P3) is automatically retrieved, Said image (P2) is automatically modified by means of image processing software (P64) and/or image restoration component (P65), taking into account said specific formatting information obtained in this way. 25. The system according to claim 24, said system enables said data processing means to automatically carry out said retrieval by means of an index (P52), said index (P52) is directly or indirectly obtained by analyzing the following content by means of analysis : the image (P2), and/or said means of said means chain (P3), and/or means for loading said image (P2) into said image processing software or component (P64, P65), and/or Means for loading an image modified by image processing software or components into an image restoration device (19). 26. System according to claim 25, said means of said means chain (P3) being identified by an identifier (60), in particular a barcode, said analysis means comprising means for determining said identifier (60) identification device. 27. System according to any one of claims 24-26, said system causing said image (P2), said index (P52) and/or said identifier (60) to be contained in the same file (P57). 28. The system according to any one of claims 24-26, said system causing said image (P2) and at least part of the specific formatting information (15) to be included in the same image file (P58 )middle. 29. A system according to any one of claims 24-26, said system further comprising storage means for storing at least part of the formatted information (15) in a database (22) in advance, said system further comprising Updating means for updating said database (22). 30. System according to any one of claims 24-26, said system enabling a device in said chain of devices (P3) to have at least one variable characteristic (66) dependent on the image (P2) , especially the focal length; a part of said specific formatted information (15) is related to a defect (P5) of a device having said variable characteristic (66), said system also comprising computing means for determining: - the value of said variable feature (66) for said image, - said part of said specific formatting information (15), by taking into account the values obtained in this way for said variable feature (66). 31. A system according to claim 30, said image being contained in a file (P57), the system being such that, for determining the value of said variable characteristic (66), said system comprises data processing means for, for example, The format of the Exif standard handles the data present in said file (P57), especially data such as the focal length. 32. System according to any one of claims 24-26, said system being such that in order to improve the quality of at least one image (P2) originating from or being sent to a device chain (P3), the The system includes data processing means (P76) for determining: a virtual device (708) exhibiting a defect (P5) comparable to at least a partial defect (P5) of at least one device of said device chain (P3) whose defect (P5 ) hereinafter referred to as the original defect (P5), virtual formatting information (709) associated with a defect (P5) of said virtual device (708), Said system is such that, in order to determine specific formatting information (15) associated with said group of devices of said device chain (P3), said data processing means (P76) comprise replacement means for converting said virtual formatted The information (709) is replaced with this specific formatting information related to said original defect (P5). 33. System according to any one of claims 24-26, said system being designed to improve the quality of at least one color plane (P20) of a color image (P21), said color plane (P20) Characterized by a designated color (P22), said specific formatting information (15) also includes data related to said designated color (P22), said system comprising computing means for using said specific formatting information (15) Said data related to color (P22) and said image (P2) to calculate a color plane (P20). 34. The system according to claim 24, said system further comprising, for said means chain (P3), a means for calculating said Computing device for unknown formatting information (15). 35. A system according to claim 34, said system being such that said computing means for computing unknown formatted information related to said chain of devices (P3) comprises means for measuring defects (P5) of said devices and and/or processing means for simulating said means. 36. The system according to claim 34 or 35, for the image capture device (1) of said device chain (P3), said system further comprising computing means for using at least one reference scene (9) in a Specific mathematical projections on the surface (10) build a synthetic image class (7) to calculate the unknown formatting information (15); the image capture device (1) captures each reference scene (9) at least one reference image (1) of which said calculation means can be used to transform said reference image (11) to correspond to said reference scene (9) by selecting from a set of parameterizable transformation models (12) to compute said unknown formatting information (15), said transformed image (13) being compared to said synthetic image class (7) With a bias (14), said unknown formatting information (15) at least partially consists of parameters of said selected parameterizable transformation model (12). 37. The system according to claim 36, said system further comprising data processing means (P76) for: calculating said deviation (14) between said transformed image (13) and said composite image class (7), Said deviation (14) is associated with said unknown formatting information (15). 38. System according to claim 36, said system enabling a device of said device chain (P3) to have at least one variable characteristic (66) dependent on the image (P2), in particular focal length and/or aperture ; a part of said specific formatting information (15) is related to a defect (P5) of a device having said variable characteristic or characteristics (66); each variable characteristic (66) can be combined with a value to form a a combination (P120) consisting of the set of said variable features (66) and said values, said system further comprising information processing means for determining said portion of said unknown formatted information (15) (P76), said determination is by means of the following operations: Select a predetermined combination (P120), For each of said predetermined combinations (120), employing the process of repeating the calculation means and data processing means (P76) according to claim 36 or 37, A process of interpolating said unknown formatting information associated with any combination (P120) is employed based on said unknown formatting information obtained at the end of said iterative process. 39. The system according to claim 34 or 35, for the image restoration device (19) of the device chain, the system further comprises a data processing device (P76), which is used to generate a characterizing image restoration device (19) ) data of defects (P5), in particular distortion characteristics, said unknown formatting information (15) at least partly consisting of said data characterizing the defects (P5) of the image restoration device (19). 40. System according to any one of claims 24-26, said system being such that said specific format associated with a device or devices of said chain of devices (P3) is determined in such a way UL information (15) so that it can be applied to similar devices (712). 41. System according to any one of claims 24-26, which system causes said image (P2) to include relevant information (P63), especially a digital signal, said system being employed in such a way that Make it retain or modify the relevant information (P63). 42. A system according to any one of claims 24-26, said system further comprising data processing means for associating information with said modified image, in particular information indicating that it has been modified. 43. The system according to any one of claims 24-26, said system being more specifically designed to improve the visual quality of said image for a viewer, in relation to the means of said chain of means (P3) Said formatted information related to a defect (P5) further comprises formatted information related to said observer's visual characteristics, in particular said observer's eye and/or brain dysfunction.

Images